Daniel Angelov @RNAstrand

When I started my first exploration and contributions to some large frequently used open source code bases, I got (1) excitement that my work can help and speed up so many developers; (2) astonished that the whole thing worked *at all* in the first place.

Amjad Masad @amasad

3 years ago

The more I know about the internet, the more am surprised anything works. Critical components are often maintained by one person, typically employed by big tech, but answers to no one, lives on a boat somewhere, and has the final word on critical decisions.

96 258 3K 688K 361

0 0 1 138 0

View Details

Daniel Angelov @RNAstrand

3 years ago

Designing robots to perform open-ended tasks in the real world is both easier than expected and harder than anticipated!

0 0 0 49 0

View Details

Daniel Angelov @RNAstrand

3 years ago

@Suhail Two pop in my mind: @SebastianThrun Probabilistic robotics and John Craig - introduction to robotics. That gives a good spectrum. Plus, checkout @rosorg !

0 0 0 223 1

View Details

Daniel Angelov @RNAstrand

3 years ago

@paulg This is similar to how I've heard story writers are using generative AI. At an important part of the story, generate 10 different future steps, cross those out and write the 11th that is much more unanticipated!

0 0 1 66 0

View Details

Daniel Angelov @RNAstrand

3 years ago

@paulg "When in doubt, optimize for interestingness." I find that to be a great strategy for any part of life!

0 0 0 90 0

View Details

Oliver Cameron @olivercameron

4 years ago

If you want a glimpse as to how every startup will build and utilize AI in ~5 years, look inside a legit self-driving car startup. It’s a front row seat to bleeding edge ML, computer vision, infra, simulation, automated training, and more, all fueled by billions of $$$.

4 4 68 0 7

View Details

Daniel Angelov @RNAstrand

4 years ago

Speaking to so many companies in the industry and looking at what testing looks for them, gave us some unique insight into the Machine Learning testing landscape!

Efemarai @efemarai

4 years ago

Excited to share our extensive overview of the current ML testing landscape! efemarai.com/post/machine-l… We hope this structured presentation of the available methods and tools will make the topic more accessible and more teams will look into improving their QA processes.

0 0 2 0 0

0 0 0 0 0

View Details

Erik Bernhardsson @bernhardsson

4 years ago

Something that old boring process literature talk about (Toyota, The Goal etc) that applies to software engineering 10000000% is that minimizing the size of the feedback loop is much more important for productivity than minimizing idleness

24 197 1K 0 134

View Details

Yann LeCun @ylecun

4 years ago

- enable machines to plan hierarchically, decomposing a complex task into subtasks. - all of this in ways that are compatible with gradient-based learning. The solution is not just around the corner. We have a number of obstacles to clear, and we don't know how. 8/N

12 19 386 0 15

View Details

Yann LeCun @ylecun

4 years ago

I believe we need to find new concepts that would allow machines to: - learn how the world works by observing like babies. - learn to predict how one can influence the world through taking actions. 6/N

13 38 534 0 27

View Details

Daniel Angelov @RNAstrand

4 years ago

@sh_reya These are great! Last but not least, make sure your models are tested **before** they are deployed! From simple data wrangling to more advanced tools like @efemarai can save a lot of headache in the long run.